Position and outcomes of Minimally Invasive transCervical Esophagectomy (MICE) in treatment of esophageal cancer

Introduction

Globally, esophageal cancer is the seventh most common cause of death due to cancer, thereby highlighting its significant health burden (1). Treatment of esophageal cancer most commonly involves (neo) adjuvant chemo(radiation) followed by esophagectomy for locally advanced tumors (2). Complications after esophagectomy are often reported in literature and can range up to 60%, with pulmonary and anastomotic complications being the most frequently observed (3-5). Literature reports pneumonia rates of 14.6% to 30% in patients after esophagectomy (6,7). These complications result in prolonged hospitalization, increased mortality, and a reduction of the quality of life (8). Esophagectomy has evolved over the years, transitioning from a transhiatal approach to transthoracic and minimally invasive techniques, with the addition of robotic assistance. These minimally invasive techniques result in good oncological outcomes, but unfortunately still entail postoperative complications. With the purpose of reducing, in particular pulmonary complications, Minimally Invasive transCervical Esophagectomy (MICE) was developed in Japan in 2015 (9). With this technique esophageal resection is performed in two stages: laparoscopic transhiatal and single-port transcervical mediastinal dissection, refer to Appendix 1 for details on the surgical technique. Since the pleural space remains intact, this may reduce the chance of postoperative pneumonia and the severity of an anastomotic leak. However, MICE carries an increased risk of injury to the recurrent laryngeal nerves (RLNs) due to the extensive bilateral upper mediastinal lymph node dissection. The aim of this review is to provide an overview of the development, outcomes thus far and the position of MICE in the current healthcare plan for esophageal cancer in Europe.

Methods

A targeted literature search was conducted in the PubMed database to identify relevant publications published between 2000 and January 2026. The search strategy incorporated Medical Subject Headings (MeSH) terms and relevant keywords related to minimally invasive transcervical esophagectomy, single-port mediastinoscopic lymphadenectomy, mediastinoscopy- and laparoscopy assisted esophagectomy, transmediastinal and/or bilateral surgical approach and robot-assisted transcervical esophagectomy. Following the initial search, titles and abstracts were screened for relevance. Articles considered relevant to the topic were subsequently assessed through full text review. Reference lists of the selected articles were screened to identify additional relevant publications. Articles were considered eligible if they were published in English and if abstract and full text were available. Only publications directly related to the topic of this article were included. Case studies, expert opinion papers, retrospective, prospective observational studies and randomized controlled trials (RCT) were eligible for inclusion. Furthermore, this study was conducted by an international author group comprising leading experts in (robot-assisted) MICE from Germany, Japan and the Netherlands.

Development and position of MICE in esophageal cancer treatment

Esophageal cancer comprises two major histological subtypes: squamous cell carcinoma (SCC) and adenocarcinoma (AC), which both show distinct geographical and racial patterns. AC is the most prevalent histological subtype for esophageal cancer in Europe, whereas SCC is the most common histological form worldwide. SCC is particularly dominant in high incidence regions such as eastern Asia, where it accounts for up to 90% of all esophageal cancer cases (2,10). SCC usually arises in the upper part of the thoracic esophagus and tends to spread extensively to mediastinal lymph nodes, particularly those adjacent to the bilateral RLNs. Consequently, comprehensive lymphadenectomy of the upper mediastinum is essential for the radicality of the esophagectomy in patients with SCC (11).

Given these disease characteristics in SCC, surgical approaches that allow for effective lymph node dissection while minimizing postoperative complications are highly desirable. Previous cervical techniques included conventional mediastinoscopy, which used specialized instruments for dissecting in the narrow mediastinal space but offered a limited view around the tip, thereby restricted its use to lymph node sampling or biopsies for staging rather than en bloc lymphadenectomy (12). With the development of MICE in Japan, it became possible to perform en bloc lymphadenectomy in the upper mediastinum along the RLN using a single-port mediastinoscopy technique, which provided an improved operative view due to the expansion of the mediastinal space by carbon dioxide insufflation. As the transthoracic phase is no longer required with MICE, the pleurae remain intact, thereby eliminating the need for single lung ventilation (9).

After the introduction of the single-port mediastinoscopic approach by Fujiwara et al. in 2015, the technique was adopted by Mori et al. [2017] shortly after (13,14). As of April 2018, the technique has been covered by the Japanese national health insurance system, thereby enhancing the likelihood of nationwide implementation.

As far as current literature shows, within Europe, MICE is only performed in selected centers in the Netherlands and Germany. In 2018, with the aim of reducing postoperative pulmonary complications after esophagectomy, MICE was introduced to the Netherlands according to the IDEAL framework. This framework describes the phases of surgical innovation: idea, development, exploration, assessment, and long-term follow-up (15). In recent years, MICE has developed into one of the standard surgical techniques for the treatment of esophageal carcinomas at Radboud University Medical Center, a Dutch tertiary referral hospital and currently the only center in the Netherlands where MICE is routinely performed as an alternative to the standard transthoracic minimally invasive esophagectomy (TMIE) procedure. Well defined criteria for selecting patients for MICE have not yet been established. The decision for MICE or TMIE is based on multiple factors, including cervical access, local extent of the tumor, tumor location (high thoracic or cervical esophageal tumors), presence of high mediastinal lymph node metastases and by taking shared decision making in consideration.

Current outcomes of MICE

As the current development shows, MICE is performed only in a limited number of specialized centers. This is primarily due to its technical complexity, the restricted and narrow working space, and the need for in depth knowledge of mediastinal anatomy from the transcervical view. As a result, only a small number of studies have been published reporting on the outcomes of MICE to date.

Fujiwara et al. [2017] reported a median number of resected thoracic lymph nodes of 37.0±11.8 in the thorax and abdomen, and a R0 resection rate of 95%. Twenty patients (33.3%) suffered vocal cord palsy confirmed by routine laryngoscopy at postoperative day 7, which remained permanent in 4 patients (6.7%). Postoperative pneumonia was observed in 6.7% of the patients and anastomotic leakage in 15% (16).

In a study of 17 patients, Mori et al. [2017] reported a median number of retrieved lymph nodes of the upper mediastinum of 10 (range, 2–23), zero cases of pneumonia and recurrent laryngeal nerve palsy (RLNP), and 5 (29%) cases of anastomotic leakage (14).

Wang et al. [2019] reported the outcomes of 80 patients who underwent single-port inflatable mediastinoscopy esophagectomy. The mean number of removed lymph nodes was 21.9±4.1. Anastomotic leakage occurred in 8.9%, anastomotic stricture in 21.25% and RLNP in 15 cases (18.8%), with dysphonia persisting in 3 patients after three months, while the other cases fully recovered (17).

A recent Chinese systematic review and meta-analysis of 9 studies (819 patients), compared the outcomes of transcervical mediastinoscopic esophagectomy to thoracoscopic assisted minimally invasive esophagectomy (18). This study found no statistically significant differences between the two techniques regarding postoperative complications, in hospital mortality, intra- and postoperative bleeding, arrhythmia, chylothorax and anastomotic fistula. In contrast, this study did find statistically significant differences between groups with shorter operative times, fewer dissected lymph nodes, a lower incidence of pulmonary complications and a shorter hospital stay in the transcervical esophagectomy group. However, the generalizability of the outcomes of this review is uncertain. Some included studies are over 10 years old and used techniques without pneumomediastinum, making them a less relevant comparison to MICE. Furthermore, limited experience with the novel technique may influence outcomes. Lastly, the review did not account for histological differences in esophageal cancer. Variations in patient characteristics, comorbidities, and neoadjuvant treatments limit the generalizability of published studies (19).

To date, only one study has addressed the outcomes of MICE in Europe. This study by Vercoulen et al. [2024] presents the first clinical outcomes of MICE from Radboud University Medical Center and reports on a prospective cohort of 75 patients, evaluating the outcomes of MICE in a Western population in the Netherlands. Outcomes showed a R0 resection rate of 96%, and a medium number of resected lymph nodes of 29 (interquartile range, 22–37). Pulmonary complications occurred in 12 patients (16%), with 4 patients (5%) developing pneumonia. Anastomotic leakage was reported in 5 patients (7%) RLNP occurred in 33 (44%) patients, confirmed by laryngoscopy. Thirty patients (91%) fully recovered from RLN palsy 1 year postoperatively (20).

Esophageal dysphagia and weight loss are common after esophagectomy. In the absence of comparative data on functional gastric conduit outcomes, it is generally assumed that conduit function is similar following minimally invasive transcervical or transthoracic approaches, suggesting comparable rates of these complications.

Data on long-term outcomes following MICE remain limited, with few studies reporting survival beyond short- to midterm follow-up. In the Dutch cohort, a two-year overall survival (OS) rate of 60% was reported. This outcome aligns with data from the National Cancer Registry (20). The study by Chen et al. [2022] compared transcervical inflatable mediastinoscopic esophagectomy with thoracoscopic esophagectomy for locally early- and intermediate-stage esophageal squamous cell carcinoma. No significant differences were found in the number of harvested lymph nodes and recurrence-free survival (RFS) rates between the two groups (21).

Adequate lymphadenectomy is generally considered as an important determinant of oncological quality in esophagectomy, with many studies reporting an association between higher lymph nodes yields and improved OS and disease-free survival in patients with esophageal cancer (22). In SCC, lymph node yields ≥24 are associated with improved OS in lymph node positive (N+) disease, whereas this benefit appears less pronounced in lymph node negative (N0) patients (23). In AC, retrieval of ≥31 lymph nodes has been linked to superior staging accuracy and improved disease free and OS without increased morbidity (24). However, not all studies report a clear relationship between the number of dissected lymph nodes and long-term outcomes (25). The surgical benefit of upper mediastinal (paratracheal) lymph node dissection in patients with distal AC remains a matter of debate (26,27). However, several leading European esophageal centers perform upper mediastinal lymphadenectomy as part of the standard surgical approach for distal AC. MICE might facilitate this more extensive dissection via an even less invasive approach. Although reported lymph node yields in the outcomes of the reported MICE studies vary, the available data suggest that oncological adequate lymphadenectomy can be achieved. However, robust comparative data and long-term follow-up are required to definitively establish the oncological equivalence of MICE to transthoracic approaches.

RLN injury

The most frequently reported complications of MICE is injury to the left RLN during lymphadenectomy. Such injury can result in RLNP, which causes dysphonia and increased the risk of pharyngeal swallowing problems, which may lead to aspiration and pneumonia after esophagectomy (28,29). Persistent RLNP may increase the risk of pneumonia later in the postoperative course. However, in the study by Vercoulen et al. [2024] the incidence of pneumonia remained low, despite a high incidence of RLNP (20). Nevertheless, it emphasizes the importance of finding strategies, such as routinely endoscopic voice and swallowing assessment, to prevent such complications. Komatsu et al. [2022] showed in a retrospective cohort study that continuous nerve monitoring (CNM) may significantly reduce the risk of RLNP and thereby improve clinical outcomes. The use of CNM enabled real time interruption of lymphadenectomy and adjustment of surgical maneuvers to prevent nerve injury. These findings support the routine use of CNM to enhance both the safety and learning curve of MICE (30). In contrast, CNM could be distracting, especially in the beginning of the learning curve, because of the high frequency of alarms. Furthermore, nerve damage can already be irreversible at time of CNM alarm, in that case the use as a preventive measure is limited. While nerve injury can range from temporary to permanent, and the range of dysphonia and pharyngeal dysphagia differ across patients, the implications for recovery, long-term functional outcomes and patient-reported quality of life are not yet fully understood. RLNP management involves collaboration between speech language therapy and otolaryngology for targeted voice and swallowing rehabilitation, adjustments to food and liquid consistency, and intracordal injection augmentation, which can be offered early after onset (31). Reinnervation techniques and thyroplasty are possible approaches for managing permanent RLNP. These treatments can make RLNP a manageable and acceptable condition.

Another key point of discussion regarding MICE is the choice between unilateral versus bilateral transcervical lymph node dissection. The traditional approach typically involves a unilateral, left sided dissection, which is well documented. However, the bilateral approach might offer advantages, such as more precise dissection of lymph nodes around the RLN by facilitating easier dissection of the right paratracheal area. While the unilateral approach is sufficient in many cases, the bilateral approach may result in higher lymph node yield, particularly for the upper third of the esophagus. However, the increased complexity of the bilateral approach comes with a higher risks of bilateral vocal cord palsy, which is a severe complication and has been identified as a major cause of tracheotomy in a 2017 study by Fujiwara et al. (16). Fujita et al. [2024] report positive outcomes in a case study with ten cases of robot assisted bilateral transcervical esophagectomy. The short-term results showed that one patient developed RLNP, and two patients experienced anastomotic leakage. No cases of postoperative pneumonia were reported, and there were no severe postoperative surgical complications (Clavien-Dindo ≥3). On average, 37.2±11.2 mediastinal lymph nodes were removed (32). The study by Daiko et al. [2024] found that bilateral mediastinoscopic-assisted esophagectomy provides comparable oncological outcomes and feasibility compared to thoracolaparoscopic esophagectomy, with significantly shorter operative time and intensive care unit stay. Despite a higher lymph node yield with the transthoracic approach, incidence of RLNP, recurrence rates (including total and locoregional recurrence), and OS did not differ significantly between groups over a median follow-up of 862 days (33).

Robot-assisted cervical esophagectomy

In 2019 the first studies were published on the use of the da Vinci single-port robot-assisted cervical esophagectomy (SP-RACE) (34,35). In Europe, SP-RACE was performed for the first time in 2024 (36). This procedure employs a single-port technique alongside a laparoscopic approach and demonstrated that total esophageal mobilization and mediastinal lymphadenectomy can be effectively performed using the da Vinci SP system. This approach has been reported to allow for the dissection of paratracheal and subcarinal lymph nodes using the transcervical access, although its reproducibility, feasibility and clinical utility requires further research (36). Benefits include improved access to high esophageal tumors, enhanced dissection of the left recurrent nerve area, and potentially fewer pulmonary complications due to extrapulmonary dissection. The use of the da Vinci SP system may be best suitable for patients with upper esophageal cancer who are not eligible for a transthoracic approach, or patient who require complementary mediastinal lymphadenectomy while preserving the esophagus after endoscopic resection of early esophageal cancer, although the latter remains experimental. The technique of single-port robot-assisted cervical esophagectomy has the potential to become a game changer in esophageal cancer surgery, with advances expected as the system and related technologies continue to evolve.

However, current single-port robotic systems have certain limitations; in patients with a small body habitus or a narrow thoracic inlet or mediastinal space, interference between the robotic arms (elbows) remains a concern. Therefore, at this stage, multi-port robotic systems such as the da Vinci Xi may still offer certain advantages and remain a valid option for robot-assisted MIE.

Currently only small-scale studies have reported outcomes of the use of robotics for transcervical mediastinal dissection during esophagectomy. Fujita et al. [2024] reported in a retrospective case study the first outcomes of robot-assisted cervical esophagectomy in which the cervical and abdominal phase of the procedure were performed simultaneously with the use of nerve integrity monitoring and without the use of muscle relaxants. This technique led to a significant reduction in total operation time and facilitates intraoperative monitoring of the RLN activity (37).

In 2025, a retrospective study of 127 patients with esophageal SCC undergoing RACE reported favorable long-term survival outcomes. Three- and five-year OS rates were 96.2% and 92.1% for cStage I, 84.8% and 82.3 for stage II and 61.8% at both time points for cStage III. Corresponding 3- and 5-year RFS rates were 94.3% and 84.3% for cStage I, 71.7% and 69.3% for cStage II, and 48.2% at both time points for cStage III. Additionally, 74 patients were paired in a propensity score matched comparison with open transthoracic esophagectomy. Outcomes showed that RACE was associated with significant better OS (38).

Nevertheless, further research is needed to evaluate the feasibility, reproducibility and the clinical effectiveness of the robot-assisted cervical esophagectomy.

Future perspectives

While existing studies provide valuable insights, high quality evidence from well-designed RCT is currently lacking. Additionally, current literature reflects early learning curve results, which should be interpreted with caution, as this may limit definitive conclusions about the relative benefits of MICE. Given its technical complexity, MICE is best performed in high volume centers, and the procedure may involve a long learning curve. While no specific data are yet available for MICE, experience from MIE suggest that learning-associated complications can occur; van Workum et al. reported that in the early adoption of MIE, 36 additional patients (10.1%) experienced learning-associated anastomotic leakage (39).

Encouraged by the initial positive findings, interest is rising within the surgical field to comprehensively evaluate the potential advantages of MICE and challenges in comparison with TMIE to determine whether MICE provide comparable clinical outcomes, including safety, efficacy, and oncological adequacy and long-term outcomes. To this end, steps are being taken to launch a RCT in Europe in the near future. Notably, a protocol for an RCT has already been published in China (40). However, this trial employs a modified variant of the MICE technique, in which esophageal mobilization is performed through a cervical incision, using extraluminal gastroscopic esophageal mobilization, without an artificial pneumomediastinum. Additionally, differences between Asian and European patient populations may limit the generalizability of these results.

Concluding remarks

To fully realize the potential benefits of MICE, it is crucial to facilitate its wider adoption beyond the pioneering centers. While MICE is already a relatively common procedure in Asia, its implementation in Europe remains in the early stages. The complexity of the technique and the specialized skills required have thus far limited its application to a select number of experienced institutions. To promote broader adoption, expanding training opportunities according to the IDEAL framework, developing standardized protocols, and strengthening collaborative networks, such as the transCervical Minimally Invasive Esophagectomy (CMIE) collaboration group, will be essential (41). This broader implementation will not only enhance the generalizability of clinical outcomes but also ensure that more patients have access to this potentially advantageous surgical treatment.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.